Smart insulin dosing software based on trend arrows is an effective means of controlling hyperglycemia in non-endocrinological inpatients with diabetes: a randomized controlled trial

Purpose Artificial intelligence and real-time continuous glucose monitoring (rtCGM) are important inventions in the history of diabetes. The aim of this study was to compare the effectiveness of two methods of glycemic control (smart insulin dosing software and traditional physician experience). Methods Subjects were randomized 1:1 into a smart insulin group (App group) and a traditional physician group (Control group). The former calculates the insulin dose based on the application (with trend arrows that respond to the magnitude and direction of glucose changes), while the latter calculates the insulin dose based on the physician's experience. All subjects underwent rtCGM and capillary blood glucose (CBG) testing during the study. The duration of the study was 5 to 7 days. We compared the mean blood glucose during the intervention. Results Baseline information was matched between the two groups. Patients in the App group had lower mean CBG (mmol/L) before dinner ( P  = 0.006) and at bedtime ( P  = 0.039) compared to the control group. The mean results of rtCGM during the study showed that patients in the App group had lower mean glucose ( P  = 0.003), glucose management index ( P  = 0.003), time in range (64.0% (51.0%, 73.0%) vs 47.0% (29.5%, 67.0%), P  = 0.001) and time above of range (35.0% (23.0%, 47.0%) vs 52.0% (29.5.0%, 70.0%), P  = 0.003). There was no difference in the incidence of hypoglycemia between the two groups of patients ( P  = 0.394). Conclusion Smart insulin dosing software with trend arrows has a better effect on controlling hyperglycemia than traditional physician experience, and does not increase the risk of hypoglycemia.